DLP and RGV laser projector

[kingbowser]

Hi, I have a NEC LT265 which has neither lamp nor color wheel. I want to replace these missing parts with a DIY RGV laser system with a TTL logic control in order to switch colors. As I am new to lasers (but I have no problems concerning electronics), I need some advices. I'll merge the three colors using the PS3 sled beamsplitters. Maybe I will add a bit of 532nm to 405nm to have a more bluish primary color.

Now my requests:
-As the DMD chip in the NEC is 0.7" wide, what can I use to expand the beam? Will two lenses, a DCX and a DCV set, with their focal points superimposed, do the job?
-What power do you suggest in order to have a good light source?
-Which diodes can I use? (As I am still an undergraduate student I don't work , so I want to make the project as cheap as possible)

Thanks, Riccardo

PS: here the projector specifications.

[Laserman532]

Hi, I have a NEC LT265 which has neither lamp nor color wheel. I want to replace these missing parts with a DIY RGV laser system with a TTL logic control in order to switch colors. As I am new to lasers (but I have no problems concerning electronics), I need some advices. I'll merge the three colors using the PS3 sled beamsplitters. Maybe I will add a bit of 532nm to 405nm to have a more bluish primary color.

Now my requests:
-As the DMD chip in the NEC is 0.7" wide, what can I use to expand the beam? Will two lenses, a DCX and a DCV set, with their focal points superimposed, do the job?
-What power do you suggest in order to have a good light source?
-Which diodes can I use? (As I am still an undergraduate student I don't work , so I want to make the project as cheap as possible)

Thanks, Riccardo

PS: here the projector specifications.

i would take each laser, launch each laser into a fiber optic cable (which scrambles the coherence) take the three output fibers bundle them as best as you can with three fibers making a triangle of diverging light, take the output and launch its output in to a rectangular light pipe. Place a single lens at the output of the light pipe (imaging its output) so that it will produce a rectangular beam of light projecting on the dmd. good luck getting the ttl to mirror flip timing correct, i built a system like that many years ago and it functioned well (but we had access to the software as a dmd development team)

[kingbowser]

thank you, this is the best post I could expect

As I'll build the system with the KES 400A (playstation 3 sled) beamsplitters, my output will be already a white light source, so I don't need to use optic cable in order to combine colors.
What if I magnify the beam before entering the multistrand optic fiber and then I shape the end of the cable directly as the rectangular shape I need?

fot the ttl I'll go with reverse engineering... sigh

[Laserman532]

thank you, this is the best post I could expect

As I'll build the system with the KES 400A (playstation 3 sled) beamsplitters, my output will be already a white light source, so I don't need to use optic cable in order to combine colors.
What if I magnify the beam before entering the multistrand optic fiber and then I shape the end of the cable directly as the rectangular shape I need?

fot the ttl I'll go with reverse engineering... sigh

you contradicted yourself. You said it is whitelight so you wont need the cables, then you asked if you magnify the beam before entering the fibers.

your only solution without using fibers is to expand the beams prior to entering the lightpipe, then you image the output of the lightpipe. I have given you the information to get you started, form here on out it is left up to experimentation. You will still have to modulate the green at a faster rate than I think you can modulate the green. Give up on the lasers and go LED's like everyone else has done.

[kingbowser]

Maybe I haven't well explained what I want to do. I want to use optic fiber to scramble laser coherence (I read the TI appnote about the subject) AND shape the beam as a rectangle (using it as a rectangular lightpipe). In order to expand the beam before entering the cable I need to enlarge its size, using one or more lenses.

For what concerns the modulation, as you can see in this photo the NEC color wheel is composed by 7 segments (2*green, 2*red, 2*blue, 1*white), and it spins at 120Hz (7200RPM).

As white and green are one after another, the green laser has to switch as a rate of about 240Hz (in fact it's a bit less than half a cycle). 240Hz is a quite low switching frequency, so there will be no problems here. Instead I think that the major problem will be the color wheel feedback system emulation, this is the part I have to reverse engineer.

I have to see if it is really a constant ratio of one seventh wheel per sector, or if the photo is tricking me

In your project which laser output power you used? I'm still hesitant about this aspect.

Sorry if I make some errors, but it's because English isn't my first language

Thanks for your support, as soon as I'll have all the materials I'll follow your experimentation advice

PS: lasers are funnier than LEDs (and they have better-looking colors) XD

[mixedgas]

I'm not trying to be mean here,

Based on years of optics experience, clean up the laser beams and combine them before the mode scrambling fibers.

As for your lens size, go look up fiber "NA" parameters.
You'll then need to calculate a ABCD matrix to get the lens size, and a thing to remember here, light is never truely parallel. That comes back to haunt you "After" you bounce off the dmds and go into the projection lens..

seriously, the processing code in the unit covers the filter's color gamut, not the laser's. So its a real crapshoot to see if you get any improvement going to laser...

Your going to be as much as 30-40% down on the green at those mod rates, without retuning the laser's PID or going AO or mechanical modulation..

It will be interesting to see if you have to "jiggle" the mode scrambling fibers to get even illumination.. I think you might..

There is a reason that the Arasor/Coherent lasers are arrays...

Words of the day,,, "Transfer Function"
Steve

[Laserman532]

true dat steve.

helpful hint - to scramble coherence speckle from fiber a "fiber jiggler" is required. All you need is a micro pager motor duct taped to the fiber add 1.5 volts...instant scrambler!

[Doc]

It will be interesting to see if you have to "jiggle" the mode scrambling fibers to get even illumination.. I think you might..

Steve

I remember reading about fiber jiggling years ago, I think it was for an endoscope illuminator source.

The system used a mini crankshaft and connecting rod setup.

[kingbowser]

I'll surely check the angles between the source and the fiber. I will combine the RGV beams before entering the fiber. Of course I'll calculate also beam divergence when looking for the lens setup.

Another reason over the laser choice is the luminous flux it can emit compared with LEDs (only phlatlight LEDs could be ok, but they are quite expensive and hard to find) and mercury-vapor lamps. <----- Wrong! ok, I received the message

The projector color ratio is digitally adjustable. I'll use that to try to set the new gamut.

In theory I'll obtain a rectangular, not coherent, low divergent RGV beam. What's wrong here? Isn't it like a focused LED (but more powerful)?

Please help me understand

Thank you for your help, I really appreciate this

PS: very interesting the "jiggling factor" XD
I'll add a little motor

[tocket]

I think it'll be rather expensive to get decent luminous power from lasers too. Blue and red will be very tricky - I don't even think a 1W Nichia 445nm LD would be sufficient with the 29 lumen it gives.

Phlatlight is probably a better route.